多种杂交算子的协同组合  

Synergic combination of multiple crossovers

作　　者：周永华[1] 林靖宇[1] 

机构地区：[1]广西大学电气工程学院,广西南宁530004

出　　处：《广西大学学报（自然科学版）》2013年第3期752-761,共10页Journal of Guangxi University（Natural Science Edition）

基　　金：广西自然科学基金资助项目(桂科自0991060)

摘　　要：为探索多种杂交算子间的相互作用对其协同效应所产生的影响及其机理,进一步验证先前对2种杂交算子协同机理分析所得到的结论能否应用于3种以上杂交算子协同的情况,以6种不同杂交算子的3～6个的各种组合做实验。首先通过实验确定每个杂交算子的收敛速度和求解质量,然后按照收敛速度快慢搭配和求解质量高低搭配的原则组合杂交算子。实验结果表明,在算子的配比合适时各组合都产生了协同效应,进一步证实了协同的机理,即组合中算子对群体多样性的调节及利用是产生协同效应的关键,同时也证实了所提多种杂交算子协同组合方法的有效性。To probe the mechanisms that the interplay among muhiple crossovers has an influence over their synergy effects and to further verify whether the conclusions draw from the analyses of the synergy mechanisms of two crossovers can be applied to the case of more than three crossovers, a number of three to six crossover combinations from six different crossovers are investigated. Firstly, the convergence rate and the solution quality of every crossover are obtained by experiment. Then, crossovers are combined on the principle that the crossovers with a quick convergence rate are mixed with the ones with a slow convergence rate, and the crossovers with a high solution quality are mixed with the ones with a low solution quality. The experimental results show that under proper proportions of crossovers all the combinations produce synergy effects. Thus, the synergy mechanisms of crossovers are verified that the key for the combination to produce synergy lies in the adjustment and usage of the population diversity by the crossovers, and the effectiveness for the proposed synergic combination of multiple crossovers is also confirmed.

关 键 词：遗传算法 杂交算子 组合 协同 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]